Seeking Advice: Best Preamp Design for Low-Noise, Balanced Microphones

[Barry Farmer]

Hi everyone,

I'm currently designing a compact, low-noise 2-channel preamp module and could use some advice on the best preamp topology and components for the job. The module needs to support both microphone and line-level inputs, have balanced XLR outputs, and feature remote volume control via a 10k linear potentiometer. Selectable phantom power (24V) is also required.

Key Design Requirements:​

• Ultra-low noise performance for professional audio applications
• Flexible power: 12-24V DC or potentially phantom power
• Balanced XLR outputs (preferably electronically balanced)
• Remote volume control using a 10k linear potentiometer with a DC supply (Similar to this https://rdlnet.com/product/st-vca3/)
• Input gain range: -60 dB to +4 dB

I’ve been considering preamp ICs like INA163, ADA4627-1, THAT1583, OPA1656, but I'm open to suggestions on what would work best for ultra-low noise performance. Additionally, I’m weighing whether to go for active balancing or a transformer-based output stage—would love some input on the trade-offs for each approach.

Has anyone here worked on something similar or have recommendations on circuit design, component choices, or best practices for achieving low-noise, high-quality audio performance? Any insights on PCB layout techniques to reduce noise and interference would also be much appreciated.

This is for a specialised system used in a church setting.

Looking forward to your thoughts!

A couple of THAT1583 with their associated remote digital preamp controllers (THAT5171 etc), would be more than adequate - even were you to achieve an anechoic chamber, in your church, the noise of people breathing, etc., would far out-weigh any electrical considerations about "noise". I believe there are "development boards" available for these IC's

 

[Cqwet Dbdfte]

Remote controlled pots are very common on the Bay, I have some, and also a pre amp with passive 7-bit Digital Step Attenuator, using relays, (128 steps using thru hole metal film resistors). It is clickfree.
As has been pointed out, OP's requirements are not very unique except maybe for the volume remote.

 

[ccaudle]

this is what I am thinking of doing

That NJM device is a single supply 5V device, which means that signals are limited to approximately +/-2V peaks relative to mid supply. That is 5 dBu max level, barely consumer grade, not suitable at all for equipment expecting professional standard levels.

 

[RCAguy]

Hi everyone,

I'm currently designing a compact, low-noise 2-channel preamp module and could use some advice on the best preamp topology and components for the job. The module needs to support both microphone and line-level inputs, have balanced XLR outputs, and feature remote volume control via a 10k linear potentiometer. Selectable phantom power (24V) is also required.

Key Design Requirements:​

• Ultra-low noise performance for professional audio applications
• Flexible power: 12-24V DC or potentially phantom power
• Balanced XLR outputs (preferably electronically balanced)
• Remote volume control using a 10k linear potentiometer with a DC supply (Similar to this https://rdlnet.com/product/st-vca3/)
• Input gain range: -60 dB to +4 dB

I’ve been considering preamp ICs like INA163, ADA4627-1, THAT1583, OPA1656, but I'm open to suggestions on what would work best for ultra-low noise performance. Additionally, I’m weighing whether to go for active balancing or a transformer-based output stage—would love some input on the trade-offs for each approach.

Has anyone here worked on something similar or have recommendations on circuit design, component choices, or best practices for achieving low-noise, high-quality audio performance? Any insights on PCB layout techniques to reduce noise and interference would also be much appreciated.

This is for a specialised system used in a church setting.

Looking forward to your thoughts!

1) "Remote volume control" may throw a monkey-wrench into possibilities. Would you be satisfied recording at 32bit? 2) By "preferably electronically balanced," you mean transformerless? Or would an impedance balanced (single-ended) output be OK, e.g. Schoeps' new CMC1?

 

[ccaudle]

"Remote volume control" may throw a monkey-wrench into possibilities.

The ThatCorp devices are made for that. They are not the cheapest, but That gives essentially a turnkey design you can use.
It is digital control though, that adds a slight additional learning curve. If you really wanted to use an analog voltage as a control you could probably whip up an Arduino sketch to read the ADC and write gain values to the That digital gain control.

 

[DaveDC]

That NJM device is a single supply 5V device, which means that signals are limited to approximately +/-2V peaks relative to mid supply. That is 5 dBu max level, barely consumer grade, not suitable at all for equipment expecting professional standard levels.

+1

It is curious that the OP is willing to put up with the ~1% THD of the NJM device. And that doesn't even include the intermodulation from the interference picked up on the control voltage cable.

This block diagram is how it would have been done 50 years ago. 25 years ago, we would have done a relay-controlled analog preamp + analog output driver with control via serial data (MIDI, RS-422, etc.). Today we would use an integrated preamp + ADC (QSC Q-SYS, etc.) sending networked audio straight into a digital console. And all at about the same price point.

 

[MidnightArrakis]

+1

This block diagram is how it would have been done 50 years ago. 25 years ago, we would have done a relay-controlled analog preamp + analog output driver with control via serial data (MIDI, RS-422, etc.). Today we would use an integrated preamp + ADC (QSC Q-SYS, etc.) sending networked audio straight into a digital console. And all at about the same price point.

[This block diagram is how it would have been done 50 years ago] -- Several years ago I worked as a "Design Consultant" with an aerospace/avionics company whose -- unofficial -- motto was....."We're the leader in 15-year old technologies"!!!

However, while such a motto does sound amusing, it is also the truth!!! When you are designing products for commercial aircraft that will be carrying hundreds of "human cargo" at a cost of -- tens-of-millions -- of dollars.....you just can't gamble on using any of the latest "whiz-bang" technologies!!! You need to use components, standards and techniques that have been -- proven -- to be completely reliable with a verified history.

A part of my job was to take PCB's that had been designed back during the 1980's and early 1990's and redesign them now using Surface-Mount Components instead of Thru-Hole Components. The amount of PCB real-estate reduction was such that we ended up being able to cram 2, 3 or 4 Thru-Hole PCBs onto a single Surface-Mount PCB!!! We even went from "double-sided" PCBs to using 4-layer and 6-layer PCBs because now, it not only "made sense", but was also more cost-effective. Multi-layer PCBs have already been proven to be reliable beyond a "shadow of a doubt", especially when fabricated by top-notch PCB-fabricators.

>> This is just a general "thrown out there" comment. That's all.....

/

 

[bigben83]

That NJM device is a single supply 5V device, which means that signals are limited to approximately +/-2V peaks relative to mid supply. That is 5 dBu max level, barely consumer grade, not suitable at all for equipment expecting professional standard levels.

Thanks @ccaudle I hadnt picked that up. back to the reasearch.

 

[bigben83]

A couple of THAT1583 with their associated remote digital preamp controllers (THAT5171 etc), would be more than adequate - even were you to achieve an anechoic chamber, in your church, the noise of people breathing, etc., would far out-weigh any electrical considerations about "noise". I believe there are "development boards" available for these IC's

Thankyou @Barry Farmer I will do some research into this.

 

[bigben83]

+1

It is curious that the OP is willing to put up with the ~1% THD of the NJM device. And that doesn't even include the intermodulation from the interference picked up on the control voltage cable.

This block diagram is how it would have been done 50 years ago. 25 years ago, we would have done a relay-controlled analog preamp + analog output driver with control via serial data (MIDI, RS-422, etc.). Today we would use an integrated preamp + ADC (QSC Q-SYS, etc.) sending networked audio straight into a digital console. And all at about the same price point.

Yes i looked at using something like the ADAU1467 to get full control. but I thought this was a bit of an overkill for what I was doing, even though I am considering a ADAU1467 board with XLR in and out for ease of use.

 

[bigben83]

I have seen there is a discussion on the THAT1512 on this forum
https://groupdiy.com/threads/that1512-pre-diy-reasonable-replacement-for-cheap-preamp-art.86210

 

[ccaudle]

Did you read the Designing Microphone Preamps presentation I linked previously from That Corp?

 

[JF]

Is it possible to bring normal AC powering to the project instead of 24V DC? How many channels do you need?

I would choose something like a RME XTC (which is in fact a THAT chip + control chip very well implemented) which you can control over Midi and build a small pot/10V to Midi controller board. I once did that for my UFX and it was way quicker done as I thought with a small arduino board where you can find Midi programs ready made for.
All together for sure cheaper as a development and you have 8 channels of great sounding preamps.

(Just detected there is no analog output at the XTC, had that wrong in mind. You would need the RME Micstasy ... way more expensive but probably still cheaper as a custom development?)

 

[Barry Farmer]

Remote controlled pots are very common on the Bay, I have some, and also a pre amp with passive 7-bit Digital Step Attenuator, using relays, (128 steps using thru hole metal film resistors). It is clickfree.
As has been pointed out, OP's requirements are not very unique except maybe for the volume remote.

I note that most of those remote volume controls feature a standard, motorised potentiometer which would be un-usable as the gain setting component of the THAT preamps. Better (by a long way) to use the THAT designed digital remote controllers.

 

[Cqwet Dbdfte]

I note that most of those remote volume controls feature a standard, motorised potentiometer which would be un-usable as the gain setting component of the THAT preamps. Better (by a long way) to use the THAT designed digital remote controllers.

A passive pot in a non-differtial circuit is very easy to implement.
There are linear motorized pots that can be used with 5170, maybe swapped in to those Bay kits.
THAT 5173 requires a microcontroller, some coding, plus some IR, Wifi or BT to a remote, with some coding.
Maybe a bit more that the typical DIY job.
Sending the design to Upwork is a bit outside the DIY spectrum.
Running those THAT chips at +/- 17V should give enough headroom.

 

[rh001]

So after some discussion this is what I am thinking of doing, just a rough block diagram.
including low high input gain jumper.
View attachment 146723

Hi!

Your design is so off target I almost don't know what to say. It's actually not possible to design something like what you want by taking the approach of ordering a la carte at a diner. Audio chips need at least +-15vdc for adequate headroom (or some discreet designs will give needed headroom with +24vdc), that NJM thing is a toy. 3vrms output is not enough, should be 10vrms at least. many semiconductor manufacturers make " high performance" this or that audio chip but in fact those chips are rarely if ever seen in any high performance equipment, only seen in consumer junk that will break and go to the landfill. I think it unlikely that mic pre-amp gain can be set-and-forget...it needs to be varied for the source at hand...so a remote gain control would be ideal in addition to remote level control but that's a pain to implement and is rarely done but yes as was mentioned in another reply that can be found in some stage boxes. Reliability is important for commercial applications and a Neve 1073 (1272) based mic pre and output stage is hard to beat as it has among other things transformer input and output. Auido toys without transformer inputs and outputs is subject to getting blown out. Cost of such a mic pre will be much less over time, even over a period of 10 years and can be repaired and will last more or less forever. If you must implement some remote stuff be sure to use reliable connectors such as xlr type (available with up to 7 (8?) pins and use high quality pots and switches. Switches should have gold contacts...try Grayhill or Elma...and pots should be hi-rel types like RV4 or Bourns series 80 or 92, or Alps pots like with form factor of RK27 or the Noble pots that are slightly smaller than RV4 I forget the series number. The commonly available affordable pots from Taiwan (Alpha) and China are for toys. Also, since for a church or any auditorium you should include high frequency rolloff pot or switch to make feedback less likely and or to suppress noises like from cell phones. Here is a good analogy for a good design...total precision as a result of grueling hard work: